Stable Diffusion · LoRA
Studio Ghibli LoRA for Stable Diffusion 2.1 (Freedom Redmond)
LoRA that applies Studio Ghibli's signature animation style to Stable Diffusion 2.1 generations using trigger words "Studio Ghibli" and "StdGBRedmAF".
About this model
This LoRA adapts Stable Diffusion 2.1 base model to produce images in the distinctive Studio Ghibli animation style — soft color palettes, expressive character designs, whimsical fantasy elements, and the hand-crafted aesthetic seen in films like Spirited Away and My Neighbor Totoro. The model was trained on the Freedom Redmond variant of SD 2.1 and uses two trigger tokens: "Studio Ghibli" for the general style and "StdGBRedmAF" as the trained instance identifier.
Recommended usage with diffusers: load the base model (stabilityai/stable-diffusion-2-1-base) in fp16 on CUDA, then load the LoRA weights from the Safetensors file. Sample prompts from the model card show it handles portraits (happy child), atmospheric scenes (ghost/horror), character designs (cat in sunglasses, pink fluffy monster, pirate boy), and fantasy compositions — all rendered in consistent Ghibli style. The custom license (bespoke-lora-trained-license) permits commercial use with rental restrictions and requires attribution.
Model details
| Category | Stable Diffusion |
|---|---|
| Subcategory | LoRA |
| Base model | stabilityai/stable-diffusion-2-1-base |
| Task | text-to-image |
| Hugging Face downloads | 76 |
| Hugging Face likes | 7 |
| Data last synced | 2026-11-12 |
Learn more
Read the in-depth guide about Studio Ghibli LoRA for Stable Diffusion 2.1 (Freedom Redmond)
Tags
diffusers · text-to-image · stable-diffusion · lora · template:sd-lora · studio ghibli · style · base_model:stabilityai/stable-diffusion-2-1-base · base_model:adapter:stabilityai/stable-diffusion-2-1-base
Explore the source
How to use it
Load the adapter with a stabilityai/stable-diffusion-2-1-base checkpoint in ComfyUI, Automatic1111 or another UI that loads LoRA files, then put the prompt in the positive prompt.